AI traditional chinese medicine initial diagnosis system

Abstract

The application discloses a traditional Chinese medicine looking, listening, asking and feeling initial diagnosis system and method based on artificial intelligence, and belongs to the technical field of traditional Chinese medicine intelligent diagnosis and treatment. The system collects looking, listening, asking and feeling information through a user interaction module, and after standardization by a data preprocessing module, inputs an AI syndrome differentiation reasoning core with dynamic optimization capability, automatically completes traditional Chinese medicine constitution identification and syndrome differentiation analysis, and finally generates an initial diagnosis report containing health suggestions and trend tracking. The application solves the problems that traditional Chinese medicine diagnosis and treatment is highly dependent on the experience of doctors, is difficult to standardize and is low in efficiency, is simple to operate, can be widely applied to daily health management and public health service scenes, and helps to build a health management system with prevention as the priority.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence and traditional Chinese medicine integration technology, specifically to an AI-powered TCM initial diagnosis system that uses a mobile phone equipped with AI applications as the core terminal to achieve high-frequency standardized collection of TCM four diagnostic methods (inspection, auscultation, inquiry, and palpation), AI intelligent processing, and syndrome differentiation. Background Technology

[0002] Traditional Chinese medicine (TCM) takes "holistic view," "syndrome differentiation and treatment," and "harmony between man and nature" as its core theories. It has unique advantages in conditioning the body, initial diagnosis of mild symptoms, and prevention of chronic diseases. However, the initial diagnosis of TCM relies heavily on the personal clinical experience of individual physicians, which leads to problems such as inconsistent diagnostic standards, uneven distribution of high-quality medical resources, and difficulty in popularizing TCM at the grassroots level. At the same time, TCM has long lacked the accumulation of large-scale, standardized, and objective clinical data, making it difficult to deeply integrate with modern technology and restricting its modernization and popularization.

[0003] In existing technologies, some TCM intelligent diagnostic solutions only collect data from a single diagnostic method, failing to achieve the overall diagnostic logic of "combining the four diagnostic methods." Moreover, most of these solutions involve single-time data acquisition, which can lead to accidental biases, and lack standardized multi-frequency data collection and intelligent data processing mechanisms. Furthermore, most solutions rely on professional medical terminals, which are complex to operate and have high barriers to entry, failing to meet the daily health management needs of ordinary users. Older mobile phones without AI applications also cannot meet these TCM intelligent diagnostic needs due to their lack of intelligent computing and data processing capabilities.

[0004] Based on the aforementioned technical pain points, this invention deeply integrates the core theories of traditional Chinese medicine with artificial intelligence, multimodal perception, and micro-sensor technology. Using widely available mobile phones equipped with AI applications as the core terminal, it achieves standardized collection of TCM diagnostic data at multiple frequencies and automatic compromise by AI. Summary of the Invention

[0005] This invention provides an AI-powered TCM initial diagnosis system that addresses the technical problems of traditional TCM initial diagnosis, such as reliance on physicians' personal experience, lack of standardized data collection, high barriers to entry for terminal devices, and lack of large-scale objective data accumulation. It enables the objectification, standardization, and popularization of TCM initial diagnosis, while providing large-scale and standardized basic data support for the verification of TCM theories and the discovery of industry patterns.

[0006] An AI-powered TCM initial diagnosis system uses a mobile phone with AI applications as its core terminal. Older phones without AI applications are incompatible with this system. The system includes a portable data acquisition hardware module, an AI core engine module, and a standardized data base module. Each module works in deep collaboration with the mobile phone with AI applications to form a complete digital closed loop of "data acquisition-processing-diagnosis-storage". The specific structure is as follows: 1. Portable data acquisition hardware module: including a wristband-style three-finger pressure-sensing pulse diagnostic instrument and a miniature bionic electronic olfactory diagnostic instrument. Both the pulse diagnostic instrument and the olfactory diagnostic instrument are equipped with positioning markers and tightness adaptation structures, supporting wireless / wired connection with mobile phones equipped with AI applications via Bluetooth / Type-C. The module utilizes the native high-definition camera of the mobile phone equipped with AI applications to form a visual diagnosis data acquisition unit and the mobile phone's own consultation unit. This module supports 3 / 5 standardized continuous acquisitions, adapting to different time nodes in the morning, noon and evening, and is used to collect pulse data, odor fingerprint data, visual data of tongue coating / face color / nail / lip color, as well as consultation data from the mobile phone AI doctor and user self-report data, and synchronizes all collected data to the mobile phone equipped with AI applications in real time.

[0007] 2. AI Core Engine Module: Deeply integrated with the dedicated AI application on mobile devices, including a multi-frequency data processing engine, a data preprocessing engine, a diagnostic model engine, and a pattern mining engine. The multi-frequency data processing engine automatically calculates a compromise value from 3 / 5 collections of data in the same dimension of the four diagnostic methods, using this compromise value as the sole diagnostic input to avoid the random bias of a single collection. The data preprocessing engine standardizes and cleans the compromise value data, extracts features, and labels it with TCM characteristic tags. The diagnostic model engine, based on the TCM theories of holistic diagnosis, Yin-Yang and Five Elements, and the correspondence between man and nature, integrates the compromise value data with the time / season / regional spatiotemporal data collected to intelligently determine the user's core constitutional imbalances, such as Yin-Yang, cold / heat, deficiency / excess, and imbalances in the Five Elements and internal organs. The pattern mining engine extracts common patterns from massive amounts of standardized diagnostic data, including symptom combinations-syndrome correspondence, the relationship between human-nature factors and syndrome types, and the interaction of internal organs, continuously optimizing the diagnostic model and updating it synchronously on mobile devices with the AI ​​application.

[0008] 3. Standardized Data Base Module: Synchronizes in real time with the dedicated AI application on the mobile device equipped with AI applications. It is used to classify and encrypt the data collected from multiple times, the compromised data after AI processing, the data of diagnosis and treatment conclusions, the basic information of users, and the spatiotemporal correlation data, so as to provide data support for AI model optimization and basic research in traditional Chinese medicine.

[0009] The mobile phone equipped with AI applications serves as the sole entry point, data relay station, and core computing terminal for the entire system. Its dedicated AI applications enable functions such as hardware connection calibration, reminders for multiple data collection time nodes, standardized AI-based consultations, synchronized data transmission, display of diagnostic conclusions, output of personalized conditioning suggestions, retention of personal health records, and tracking of constitution trends. The entire process requires no professional TCM knowledge or operational skills from the user.

[0010] This invention also discloses the application process of the aforementioned AI-based TCM initial diagnosis system, including the following steps: Step 1: Connect the portable data acquisition hardware module to a mobile phone with an AI application installed via Bluetooth / Type-C, open the dedicated AI application on the mobile phone to complete the device calibration, select the 3-time / 5-time standardized data acquisition rule, and the mobile phone will automatically generate reminders for the morning, noon and evening data acquisition time nodes; Step 2: Users follow the time reminders on their mobile phones to complete the standardized collection of data from 3 / 5 observation, auscultation, palpation, and AI-standardized consultations, obtaining structured symptom data. All raw collected data is synchronized in real time to mobile phones with AI applications installed and uploaded to the AI ​​core engine module. Step 3: The AI ​​core engine module automatically calculates the compromise value of data in the same dimension through the multi-frequency data processing engine. After the data preprocessing engine cleans and labels the data, the dialectical model engine completes the intelligent fusion and dialectical analysis, which takes ≤10 seconds. The results of the analysis are synchronized to the mobile phone with the AI ​​application installed in real time. Step 4: The mobile app with AI installed displays the initial diagnosis of the user's core constitution imbalance and, based on the core constitution imbalance, outputs personalized conditioning suggestions for diet, rest, emotions, and simple health maintenance that are relevant to daily life. Step 5: After anonymizing the original data from multiple initial diagnoses, the AI-compromised data, and the diagnostic conclusions, the data is synchronized to the personal health record and standardized data base module on the mobile phone equipped with the AI ​​application, completing a standardized initial diagnosis and data collection loop, and then input into the personal health record storage.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention uses a mobile phone equipped with AI applications as the core carrier, abandons professional medical terminals, greatly reduces the threshold for using TCM initial diagnosis, realizes "standardized TCM initial diagnosis anytime and anywhere", promotes the popularization of TCM to ordinary families and grassroots medical scenarios, and clarifies that old mobile phones without AI applications cannot be adapted, thus defining the hardware foundation and scope of use of the system. 2. Innovatively proposes a data processing rule of standardized collection of data at multiple frequencies of 3 / 5 times + AI automatic median value taking, which avoids the random bias of single collection from the source of data, improves the objectivity and accuracy of diagnostic data, and requires no user intervention throughout the process, providing a brand-new solution for the objective collection of TCM four diagnostic data; 3. To realize the digital implementation of the four diagnostic methods of TCM, namely observation, auscultation, inquiry and palpation, and to integrate the core theories of Yin-Yang and Five Elements and the correspondence between man and nature, so as to transform the vague diagnostic system of traditional Chinese medicine into a quantifiable and replicable intelligent diagnostic system, thereby breaking the dependence of traditional Chinese medicine's initial diagnosis on the personal experience of physicians. 4. Constructing a positive cycle of "application-collection-optimization" ensures that each standardized initial diagnosis accumulates a high-quality basic data for traditional Chinese medicine, filling the long-standing gap in the industry of traditional Chinese medicine lacking large-scale, standardized, and objective data, and providing data support for the verification of traditional Chinese medicine theories, the study of the mechanism of organ linkage, and the exploration of the law of correspondence between man and nature; 5. The AI ​​core engine module has dynamic self-optimization capabilities. The pattern mining engine can continuously mine common patterns from massive amounts of data, constantly optimize the dialectical model, and achieve "the more it is used, the more accurate the diagnosis", thereby enhancing the long-term value of the system. 6. The system is simplified throughout the entire process, which can be completed independently by ordinary users. The time consumption of each process is flexible and adaptable. It also has the functions of retaining personal health records and tracking physical fitness trends, which can meet the daily health management needs of ordinary users and help build a public health management system based on prevention. Attached Figure Description

[0012] Figure 1 This is a flowchart of the method of the present invention and a schematic diagram of the overall architecture of the AI ​​TCM initial diagnosis system. (1) Wristband-type three-finger pressure sensor pulse diagnosis device: corresponding to traditional Chinese medicine palpation, simulating a doctor's "pulse diagnosis", collecting data such as pulse frequency, pressure, and rhythm.

[0013] (2) Mobile phone observation and collection unit: corresponding to traditional Chinese medicine observation and collection, use the mobile phone camera to take pictures of the tongue and face color, and AI automatically analyzes the characteristics of tongue color, tongue coating and face color.

[0014] (3) Miniature bionic electronic nasal auscultation device: corresponding to traditional Chinese medicine auscultation, it collects odor data such as breath and body odor through sensors to help determine the cold and heat and deficiency and excess in the body.

[0015] (4) AI Standard Consultation Multimedia Cockpit: Corresponding to TCM consultation, it is essentially a standardized consultation interaction interface. Users answer questions about symptoms, medical history, diet and sleep in their mobile phones / cabins according to AI guidance. The system automatically records the answers, replacing the traditional manual consultation.

[0016] (5) Portable data acquisition hardware module: It is the unified data entry point for all front-end acquisition devices (pulse diagnosis, inspection, auscultation and olfaction, inquiry diagnosis), and it aggregates and transmits the data from different devices to the system.

[0017] (6) Data preprocessing engine: Cleans the collected raw data: removes noise, unifies the format, completes the data, and converts non-standard raw data into a format that the AI ​​model can recognize.

[0018] (7) Mobile core board with AI application: The computing and control center of the entire system. All AI models run locally on this mobile core board to realize edge inference without relying on the cloud.

[0019] (8) AI Diagnosis Model Engine: (1st) Primary Diagnosis Model: Based on the preprocessed four diagnostic data, quickly output the preliminary TCM syndrome judgment (such as "wind-cold common cold" and "liver stagnation and spleen deficiency") as the initial screening result.

[0020] (9) AI Diagnosis Model Engine: (2nd one) Advanced Diagnosis Model: Based on the initial results, it combines more complex multimodal features (such as cross-validation of pulse + tongue + consultation data) to output more accurate diagnosis conclusions and treatment suggestions. There are obvious advantages to having two models: 1. Clear division of labor: The initial screening model is responsible for fast and low-computing-power processing, while the advanced model is responsible for in-depth analysis, reducing the computing power pressure on the end side; 2. Strong maintainability: The two models can be iterated and optimized independently without affecting the overall process; 3. Clear process: It conforms to the diagnosis and treatment logic of traditional Chinese medicine of "first screening, then detailed differentiation".

[0021] (10) Multimodal acquisition module: Integrates four different types of data: pulse, observation, auscultation and inquiry, to achieve cross-modal fusion analysis and avoid the bias of single data.

[0022] (11) Multi-frequency acquisition module: Supports repeated acquisition at different times (such as comparing the tongue / pulse of the same user on different dates) to dynamically track health changes, rather than making a single static judgment.

[0023] (12) Standardized database module: The AI ​​diagnosis results and user data are uniformly standardized and archived according to TCM syndrome type, symptoms, constitution and other dimensions to facilitate subsequent query, comparison and model training.

[0024] (13) User First Visit Outpatient Archive: Stores each user's personal health record, including data from the four diagnostic methods collected each time, diagnosis results, treatment suggestions, etc., forming a complete personal health closed loop, which facilitates subsequent follow-up visits for comparison and long-term health management.

[0025] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. Example 1

[0026] An AI-powered TCM initial diagnosis system uses a smartphone with AI applications as its core terminal. Older, non-smartphones without AI applications are not compatible with this system. The specific settings of each module are as follows: 1. Portable data acquisition hardware module: The wristband-style three-finger pressure sensor pulse diagnostic instrument is equipped with cun, guan, and chi positioning marks and an elastic band, supports Bluetooth 5.0 for wireless connection with mobile phones, and collects data such as pulse depth, slowness, rate, wiryness, and strength; the miniature bionic electronic nose is equipped with a portable buckle, supports Type-C direct connection with mobile phones, and collects odor fingerprint data of breath / body odor; relying on the 48-megapixel high-definition front camera of the mobile phone to build a visual diagnosis data acquisition unit, capture high-definition images of tongue coating, complexion, nails, and lip color, as well as standardized mobile AI consultation and user self-report; this module supports standardized continuous acquisition 3 times / day, adapting to three time nodes: 7:00 am, 12:00 pm, and 7:00 pm.

[0027] 2. AI Core Engine Module: Deeply integrated with the dedicated AI application on mobile devices, the multi-frequency data processing engine completes calculations on the mobile device, taking the arithmetic mean of the four diagnostic methods data collected from three different sources as a compromise value; the data preprocessing engine performs noise reduction and feature extraction on the compromise value data, and labels it with TCM characteristic tags such as "white and greasy tongue coating," "slow pulse," and "sour and foul breath"; the syndrome differentiation model engine integrates the compromise value data with the time and season data at the time of collection, and completes intelligent syndrome differentiation based on the model trained on classic medical cases such as "Huangdi Neijing" and "Shanghan Lun"; the pattern mining engine uses cloud computing to mine the correlation patterns between symptom combinations and TCM syndrome types.

[0028] 3. Standardized Data Base Module: It adopts cloud-based encrypted storage and synchronizes in real time with the dedicated AI application on the mobile device. The data is managed hierarchically according to the data types of "inspection / smell test / inquiry / palpation", the syndrome types of "cold / heat / deficiency / yin / yang", and the scenario dimensions of "time / season". All data is anonymized and only the syndrome-related characteristics are retained to protect user privacy.

[0029] The application process of the AI-based TCM initial diagnosis system in this embodiment is as follows: 1. Open the dedicated AI application on your mobile phone, connect the wristband pulse diagnostic device to your phone via Bluetooth, and connect the mini electronic nose directly to your phone via Type-C. After completing the device calibration, select 3 standardized data collection rules, and the mobile phone will automatically generate collection time reminders at 7:00, 12:00, and 19:00. 2. Users complete the following three time points according to the time reminder on their mobile phones: inspection (taking high-definition photos of tongue coating / facial complexion), auscultation (collecting breath data with an electronic nose), palpation (collecting pulse data by wearing a pulse diagnosis device), and AI-standardized consultation and answering 10 basic questions about diet, sleep, bowel movements, etc., generating structured symptom data. All raw data is synchronized to the mobile phone in real time. 3. The mobile AI application automatically calculates the arithmetic mean of the three collected data points. After being cleaned and labeled by the preprocessing engine, the data is uploaded to the syndrome differentiation model engine. The engine completes intelligent syndrome differentiation within 10 seconds and determines the direction of the user's core constitution imbalance of "spleen deficiency and phlegm dampness". 4. The mobile app displays the initial diagnosis of "spleen deficiency and phlegm dampness" and, based on the imbalance of this constitution, provides personalized conditioning suggestions in terms of diet, rest, emotions, and simple health preservation. 5. The original data collected from the three initial consultations, the AI-adjusted data, and the diagnostic conclusions were anonymized and then synchronized to the personal health record on the mobile device and the standardized data base module in the cloud, completing a standardized initial consultation and data collection loop.

[0030] Example 2

[0031] The core difference between this embodiment and Embodiment 1 is: 1. The portable data acquisition hardware module supports 5 standardized continuous data acquisitions, adapting to four time points: morning, noon, evening, and bedtime (including 2 data acquisitions at night). 2. The multi-frequency data processing engine completes the calculation in the cloud, taking the weighted average of the four diagnostic methods data collected from five times in the same dimension as a compromise value, and giving higher weight to the data collected from the time of the TCM organs. 3. The diagnostic model engine, in addition to integrating compromise value data, time of day, and seasonal data, also integrates regional data (such as the hot and humid south and the dry north) to further improve the accuracy of diagnosis. 4. On the mobile app, based on the user's core constitution imbalance of "liver stagnation turning into fire", personalized conditioning suggestions on diet, rest, emotions, and simple health preservation are provided according to their constitution.

[0032] The parts of this invention not described in detail are prior art. The scope of protection of this invention is determined by the claims. Any equivalent modifications based on this invention are within the scope of protection of this invention.

Claims

1. An AI-powered TCM initial diagnosis system, characterized in that, With mobile phones equipped with AI applications as the core terminal, older mobile phones without AI applications cannot be adapted to this system. The system includes a portable data acquisition hardware module, an AI core engine module, and a standardized data base module. Each module is deeply integrated with the mobile phone equipped with AI applications to form a closed-loop digital process. The portable data acquisition hardware module includes a wristband-style three-finger pressure-sensing pulse diagnostic instrument and a miniature bionic electronic olfactory diagnostic instrument. It forms a visual diagnostic data acquisition unit based on the native high-definition camera of a mobile phone equipped with AI applications, and a consultation unit utilizing the mobile phone's own functions. The pulse diagnostic instrument and the olfactory diagnostic instrument are equipped with positioning marks and tightness adaptation structures, and support Bluetooth / Type-C connection with mobile phones equipped with AI applications. The portable data acquisition hardware module supports 3 / 5 standardized continuous acquisitions, used to collect pulse data, olfactory fingerprint data, visual data, and consultation data, and synchronize them to the mobile phone equipped with AI applications. The AI ​​core engine module works in conjunction with a dedicated AI application on a mobile device, including a multi-frequency data processing engine, a data preprocessing engine, a diagnostic model engine, and a pattern mining engine. The multi-frequency data processing engine automatically calculates a compromise value for 3 / 5 times of data collected in the same dimension, and uses this compromise value as the sole diagnostic input. The diagnostic model engine is based on the four diagnostic methods of traditional Chinese medicine, the theory of Yin and Yang and the Five Elements, and the theory of the correspondence between man and nature. It integrates the compromise value data with the time / season / regional spatiotemporal data at the time of collection to intelligently determine the direction of the user's core constitution imbalance. The standardized data base module is synchronized in real time with the dedicated AI application on the mobile phone equipped with AI applications. It is used for hierarchical, classified and encrypted storage of frequently collected raw data, AI compromise data, dialectical conclusion data, user basic information and spatiotemporal correlation data. The mobile phone equipped with AI applications serves as the system's sole operation entry point, data relay station, and core computing terminal. Its dedicated AI applications enable hardware calibration, data collection time reminders, standardized AI-based diagnosis, display of diagnostic conclusions, personalized conditioning suggestions based on core constitution imbalances, and the retention of personal health records.

2. The AI-based TCM initial diagnosis system according to claim 1, characterized in that, The multi-frequency data processing engine can perform calculations on mobile devices or in the cloud with AI applications installed. The compromise value is the arithmetic mean or weighted average of the data collected in the same dimension, and higher weight is given to the data collected from the time of the TCM organs.

3. The AI-based TCM initial diagnosis system according to claim 1, characterized in that, The core imbalances in the constitution determined by the diagnostic model engine include Yin and Yang, cold and heat, deficiency and excess, and imbalances in the five elements and internal organs. The initial diagnosis conclusion output by the engine is a concise conclusion based on non-professional subdivision of the syndrome.

4. The AI-based TCM initial diagnosis system according to claim 1, characterized in that, The portable data acquisition hardware module performs 3 / 5 standardized continuous data acquisitions to adapt to different time nodes in the morning, noon, and evening. During the data acquisition process, a dedicated AI application on a mobile phone with AI installed generates intelligent time reminders.

5. The AI-based TCM initial diagnosis system according to claim 1, characterized in that, The data stored in the standardized data base module is anonymized, retaining only the syndrome differentiation-related features to protect user privacy and provide data support for basic research in traditional Chinese medicine.

6. The application process of the AI-based TCM initial diagnosis system according to any one of claims 1-5, characterized in that, Includes the following steps: Step 1: Connect the portable data acquisition hardware module to a mobile phone with an AI application installed via Bluetooth / Type-C, open the dedicated AI application on the mobile phone to complete the device calibration, select the 3-time / 5-time standardized data acquisition rule, and generate a data acquisition time node reminder on the mobile phone. Step 2: Users complete 3 / 5 observation, auscultation, and palpation data collections according to the time reminder on their mobile phones, and complete 1 AI standardized consultation on their mobile phones. All raw collected data are synchronized to mobile phones with AI applications installed and uploaded to the AI ​​core engine module. Step 3: The AI ​​core engine module calculates the compromise value of data in the same dimension through the multi-frequency data processing engine. After the data preprocessing engine cleans, extracts features, and labels TCM tags, the syndrome differentiation model engine completes the intelligent syndrome differentiation, which takes ≤10 seconds. The syndrome differentiation results are synchronized to the mobile phone with the AI ​​application installed. Step 4: The mobile app with AI installed displays the initial diagnosis of the user's core constitution imbalance and, based on the core constitution imbalance, outputs personalized conditioning suggestions for diet, rest, emotions, and simple health maintenance that are relevant to daily life. Step 5: After anonymizing the original data from multiple initial diagnoses, the AI-compromised data, and the diagnostic conclusions, the data is synchronized to the personal health record and standardized data base module on the mobile phone equipped with the AI ​​application, completing a standardized initial diagnosis and data collection loop.

7. The application process according to claim 6, characterized in that, The dialectical model engine can dynamically optimize the model by mining common patterns from massive standardized data through a pattern mining engine. The optimized model is then updated in real time to mobile devices with AI applications installed.

8. The application process according to claim 6, characterized in that, The personalized conditioning suggestions in step 4 are tailored to the daily operational needs of ordinary users. They are generated specifically based on the user's core physical imbalance and do not involve restrictions on specific ingredients or medicinal materials.

Images